特曼：史丹福_矽谷的產學合作如何創造雙贏？

林健正寫於 2012年9月3日 1:33

學術與產業兩者之間的關係存在著簡單的邏輯，學生畢業有好的就業機會，學校就容易招到好的學生，學校就容易產出好的教學及研究成果，優秀的學生畢業後投入產業，產業就會有較好的發展條件，企業會有更多的資源回饋學校。政府為了發展經濟，增加就業機會，也會在這些領域錦上添花挹注更多的資源。

特曼教授（Professor Frederick Terman)被人稱之為「矽谷之父」，也是史丹福大學邁向頂尖的推手。在1930年代，他任教於該校電機系時，就深刻地體會產學合作創造雙贏的關鍵，這個道理則悟自他在MIT的指導教授Professor Vennevar Bush，這位曾經擔任羅斯福總統科技顧問的教授與MIT附近的國防科技公司有相當密切而且頻繁的互動。

在第二世界大戰以前，史丹福大學附近仍是柑橘園，沒有什麼產業，學生畢業就到東海岸找工作。史丹福工學院之所以能夠在全球學院的競爭中領先，與特曼教授做為學術帥才的獨到眼光十分有關。當時，特曼教授認為先要以增加就業機會，設法留住人材。因此，學校應該幫助企業界發展。

他鼓勵工學院的同仁擔任企業界的顧問，教授帶著研究生做產學計畫，學校為企業開設榮譽課程，他甚至騰出空間邀請企業在校園內成立實驗室。他協助包括Willian Hewlett與 David Parkard兩位學生創業，他借了538元給這兩位研究生，出面背書保證，幫忙辦理銀行貸款等等，惠普就是在這樣的時空背景開始。他深信企業發展地好，學校就會得到好處。

特曼教授認為史丹福大學要晉陞成為頂尖的大學必須用高薪聘用一流的教授。如果有6萬元可以聘人，以當時的待遇，一般人會用來聘4個人，特曼教授卻不這麼做，他用其中3萬元聘2位年輕的助理教授，再以另外3萬元聘請另外一個資深傑出教授。結果沒多久之後，史丹福大學就出了一位諾貝爾獎得主。有了諾貝爾獎得主之後，史丹褔大學市就更能吸引更多的學生和學者，匯集更多優秀的人才。

到了1950年代，特曼教授建議學校董事會，在史丹福大學校園內成立史丹福工業區(Stanford Industry Park，後來更為Stanford Research Park)，他當時主要考量在於增加校務基金的收入以及方便學生找工作，卻產生意想不到的成果。企業為了貼近學校，紛紛進駐史丹福工業區，有些大企業包括奇異、惠普、尹士曼柯達、洛克希德等也把研究部門設在這裡，形成完整的產學研聚落，奠定高科技產業發展的有利條件，它的外溢效應就是矽谷的誕生，特曼教授因而得到「矽谷之父」的美名。

特曼教授在擔任史丹福大學教務長及副校長期間，他發現史丹福在於MIT競爭國防科技研究計畫常居於劣勢，於是他做了一件非常重要的決定，把相關領域的師生規模擴大一倍，用人海戰術提升史丹福大學的總體競爭力。受惠於這些國防科技研究計畫以及豐富的專利權利金收入，如今史丹福大學與MIT在美國東西兩岸，分庭抗禮，執全球高等教育與學術研究之牛耳。史丹福與矽谷建構了柏克 萊加大資訊學院院長薩克瑟尼安(Professor AnnaLee Saxenian)所稱之「專業網絡型的產學研合作體系」，特曼教授一生功不可沒。

到了他的晚年，有人問特曼教授，他期待史丹福是教學單位，還是研究單位，他毫不猶豫地回答，史丹福大學是一個知識交流與學習的場域。我國籌設竹科時，聘特曼教授為顧問，他也曾是南韓政府的座上賓，著名的「特曼報告書(Terman Report)」決定南韓高科技產業的策略及基本方向。這些都是特曼教授留下來的珍貴資產，產學合作可以創造雙贏，史丹福大學-矽谷的成功經驗說明得再清楚不過了。

特曼：產學合作要怎麼合作?

林健正寫於 2012年9月4日 2:28 ·

一九五O年代，特曼教授鼓吹史丹福大學成立史丹福研究院(Stanford Research Institute, SRI)，積極推動以實用考量做科學研究，雖然這種做法和傳統大學的角色並不相容。

其次透過榮譽合作課程(Honors Cooperative Program)，史丹福大學對企業人士敞開大門，鼓勵業界工程師直接申請就讀研究所課程，同時將史丹福大學的課程帶進企業內部。

特曼教授致力推動成立的史丹福工業園區(Stanford Industry Park)，則是全美第一個大學的工業園區，把土地租給對史丹福大學發展有利的高科技公司，結果企業自然而然地聘請該校的教授擔任顧問，許多工程師也來自該校畢業生，這些企業也會介入大學裡攸關他們事業發展的各項研究計畫，最後學校得到好處。

類比於史丹福_矽谷模式，竹科顯然受惠於交大校友創業的精神，交大也因竹科而有今天的成就，交大與竹科如魚得水，產學合作是交大的傳統核心價值，矽谷人才回流亦因竹科的發展，若台灣經濟發展遲滯人才自然而然就會外流。

然而，無論什麼時候，什麼地方，校園內總有一股純學院派的思維，認為產學合作會影響學術發展。教育部的五年五百億(頂尖大學)計畫也無可避免地產生一些後遺症，而其中之為大就在於學術與產業的疏離，包括大幅度地縮減在職專班的招生名額，靠著爭食來自教育部的資源，即可養尊處優，辛苦產學合作，何苦來哉?發表一些縱使算是無關宏旨的學術論文，亦可沽名釣譽，又何樂不為?

謝依君、謝宏仁：台灣產業升級的迷思～從中國過渡到世界市場

謝依君、謝宏仁：台灣產業升級的迷思～從中國過渡到世界市場_

Taiwan’s Trade Potential: Risks and Opportunities in a Changing Global Trade Context

Taiwan’s Trade Potential:

Risks and Opportunities

in a Changing Global Trade Context

Peter C.Y. Chow, City University of New York &

National Taiwan University

Dan Ciuriak, Department of Foreign Affairs and International Trade, Government of Canada

Abstract

In recent years, global growth in trade in goods and services has been boosted by so-called “integrative trade” in which an increasing share of traded goods and services consists of inputs into production processes, as these are being fragmented with elements distributed world-wide (“off-shoring”). At the same time, in the context of stalled multilateral trade negotiations, major trading economies are seeking free trade agreements (FTAs) to secure their market access objectives, creating risk of trade diversion for third parties.  This paper will use simulations with the GTAP computable general equilibrium model and the TradeSim gravity model of international trade, and, to examine Taiwan’s actual trade and trade potential by trade partner and by commodity, with particular attention risks posed by exclusion from East Asia’s emerging set of FTAs and the opportunities from concluding an FTA with the United States.

I.     Introduction

In recent years, much of the impetus to global growth in trade in goods and services has come from so-called “integrative trade” in which an increasing share of traded goods and services consists of inputs into production processes, as these are being fragmented with elements distributed world-wide (“off-shoring”). There are several main drivers for this trend, including the ongoing reduction in the cost of information management and communication which permits the management of production on a geographically widely distributed basis; improved logistics for shipping components; and much remaining scope for wage arbitrage as large population developing economies increase their degree of integration into the global economy.

At the same time, in the context of stalled multilateral trade negotiations, major trading economies are seeking free trade agreements (FTAs) to secure their market access objectives. By the end of 2007, total FTAs registered at the WTO have exceeded 300. With few exceptions, most WTO members signed at least one FTAs with other members.

The combination of these two trends creates both opportunities and risks for trading economies such as Taiwan—opportunities to take better advantage of domestic capabilities to acquire specialized tasks that are becoming newly contestable in the global division of labor; and risks of being bypassed in this process due to disadvantageous market access prospects because of exclusion from regional FTAs.

At first blush, the risk of serious erosion of Taiwan in international trade seems modest: since acceding to the World Trade Organization (WTO) at the beginning of 2002, Taiwan’s two-way trade in goods and services has risen much faster than GDP, such that two-way trade has increased from 95.4% of GDP (2001) to 134.3% (2006)[1].  However, in the process, Taiwan has become more dependent on Asia for export markets, with the share of merchandise exports going to Asian destination having increased from 53.2% in 2001 to 65.4% in 2006[2].  Moreover, East Asian regionalism is at an early stage as yet, at least in terms of implementation of free trade commitments.

This paper examines Taiwan’s actual trade and trade potential by trade partner and by commodity sectors, with particular attention to risks posed by exclusion from East Asia’s emerging set of FTAs and the opportunities from concluding an FTA with the United States.  Amongst the many possible scenarios, we focus on a comprehensive FTA involving the major East Asian economies, Taiwan’s major trading partners and competitors.

The paper is organized as follows.  Section 2 estimates, using the Global Trade Analysis Project (GTAP) computable general equilibrium (CGE) model, the impact on Taiwan of full implementation of an East Asian FTA.  Section 3 examines the sectoral implications of the trade diversion identified in Section 2 and considers the opportunities in third markets, based on simulations with the TradeSim gravity model of international trade.  Section 4 considers the scope for offsetting any trade losses due to preferential trade agreements struck by Taiwan’s trading partners through an FTA of its own with the United States. Section 5 draws some conclusions for Taiwan’s trade policy.

II. The Impact of East Asian FTAs on Taiwan’s Trade

East Asia has become a hotbed for bilateral and regional trade negotiations.  All the major economies in the region are in pursuit of preferential market access both with other regional partners and outside the region.  Taiwan is no exception.  It has signed individual deals with Panama, Guatemala and Nicaragua as well as a trilateral deal reached in May 2007 with El Salvador and Honduras[3], continues to pursue a deal with the United States and has put out feelers with the European Union[4].  However, given the political issues raised for Taiwan’s trading partners when it comes to FTAs, Taiwan is at a decided disadvantage in gaining access to such agreements, not only with its major regional trading partners but also with the major global economies.

Not surprisingly, this situation has led many observers to worry about Taiwan’s marginalization as a trading economy.  How significant such concerns should be is an open question.  On the one hand, FTAs have the potential to divert trade, imposing economic welfare costs on excluded parties.  On the other hand, FTAs can also boost economic activity in the partner economies, generating additional demand, which could partly or fully offset the trade lost by third parties due to the trade diversion effect.  In short, it is possible that the income effect could offset the price effect.

 Several studies on the impacts of East Asian economic integration on Taiwan’s economy were based on various scenarios on the economic integration such as ASEAN plus China, ASEAN plus 3 ( China, Japan and Korea ). [5]Yet, the development of Close Economic Partnership Agreement between China and Hong Kong (CEPA-CHK) made it necessary to consider China and Hong Kong as an integrated economic unit. Moreover, Taiwan’s trade with China and Hong Kong has had an increasing trend with 34.3% of its exports destined to CEPA-CHK in 2002-2003. Meanwhile, though ASEAN countries have 10 members, in terms of Taiwan’s trade with and investment in the region, the most significant components are its original 5 members plus Vietnam. The ASEAN-6, which include Indonesia, Malaysia, the Philippines, Singapore, Thailand and Vietnam accounted for 11.8% of Taiwan’s exports in the same period.

To shed some quantitative light on this issue, we examine the impact on Taiwan of a comprehensive FTA involving the major East Asian economies – Japan, China, South Korea, Hong Kong, and the ASEAN 6. Taiwan’s export to these 10 economies

in the region accounted for more than three quarters of its total exports in 2002-2003.

Technical background and caveats

To simulate the impact of an East Asian FTA which excludes Taiwan on the Taiwanese economy, we use the Global Trade Analysis Project (GTAP) computable general equilibrium (CGE) model, version 6.0.[6] This model integrates data on bilateral trade flows, trade protection and domestic support together with input-output tables that describe the internal economic linkages within each economy. This allows the model to generate estimates of the impact of preferential tariff elimination under FTAs on trade flows, economic output, employment and economic welfare.

The simulations are conducted with the global economy disaggregated into 26 regions (see Table 6 below for the full list). The base year for the GTAP 6.0 data is 2001; in other words, the model depicts the global economy as it was in 2001, including the size of trade flows, the level of protection and support for trade in the various economies, as well as the size and composition of GDP and other economic variables for each country/region. The protection data in the GTAP 6.0 database come from Market Access Map (MAcMap), which was produced and is maintained collaboratively by the Paris-based Centre d’Etudes Prospectives et d’Informations Internationales (CEPII) and the International Trade Centre (ITC) in Geneva. The tariff data are compiled at the Harmonized Tariff System 6-digit level and include the ad valorem equivalent of specific tariffs and the tariff equivalent of tariff rate quotas (TRQs). The GTAP 6.0 protection data are  current only as of 2001; accordingly, we update these to take into account the full implementation of the Uruguay Round tariff cuts, China’s accession commitments to the WTO, and the expiry of the WTO Agreement on Textiles and Clothing (ATC)[7]. The simulations reported below involve full elimination of trade protection as captured in the GTAP database, updated as described above, for all industrial and agricultural sectors. 

We do not attempt to model services or investment liberalization for several related reasons:

(a)   reliable information on the level of protection provided by the various domestic measures affecting services trade and foreign direct investment is lacking;

(b)   estimates of the potential response of services trade and investment flows to changes in regulatory barriers are crude at best;

(c)   liberalization of services trade and/or investment in FTAs tends to be specific to particular sectors and it is impossible to know ex ante which sectors in which economies might be subject to liberalization.

The simulations were conducted on a fully disaggregated sectoral basis (57 sectors, of which 43 are merchandise). The standard GTAP 6.0 parameter set was used; the key Armington parameters (the elasticities of substitution between products according to country of origin) have recently been updated based on new econometric research. These elasticities are on average lower than those used in some other models such as the World Bank’s Linkage model; the estimated trade and welfare impacts reported here are thus relatively conservative.[8] The definitions of the GTAP merchandise trade sectors are given in Table 1 below, along with the values of the Armington elasticities of substitution.

Table 1: GTAP sectors and Armington elasticities of substitution

Manufacturing Sectors	Armington Elasticities
Full GTAP description	Domestic vs. Imports	Between alternative sources of imports
Paddy rice	5.1	10.1
Wheat	4.4	8.9
Cereal grains nec	1.3	2.6
Vegetables, fruit, nuts	1.9	3.7
Oil seeds	2.5	4.9
Sugar cane & sugar beet	2.7	5.4
Plant-based fibres	2.5	5.0
Crops nec	3.3	6.5
Cattle, sheep, goats ,horses	2.0	4.0
Animal products nec	1.3	2.6
Wool, silk-worm cocoons	6.4	12.9
Forestry	2.5	5.0
Fishing	1.3	2.5
Coal	3.0	6.1
Oil	5.2	10.4
Gas	17.2	34.4
Minerals nec	0.9	1.8
Meat: cattle, sheep, goats, horse	3.8	7.7
Meat products nec	4.4	8.8
Vegetable oils & fats	3.3	6.6
Dairy products	3.7	7.3
Processed rice	2.6	5.2
Sugar	2.7	5.4
Food products nec	2.0	4.0
Beverages & tobacco products	1.1	2.3
Textiles	3.8	7.5
Wearing apparel	3.7	7.4
Leather products	4.1	8.1
Wood products	3.4	6.8
Paper products & publishing	3.0	5.9
Petroleum & coal products	2.1	4.2
Chemical, rubber, plastic products	3.3	6.6
Mineral products nec	2.9	5.8
Ferrous metals	3.0	5.9
Metals nec	4.2	8.4
Metal products	3.8	7.5
Motor vehicles & parts	2.8	5.6
Transport equipment nec	4.3	8.6
Electronic equipment	4.4	8.8
Machinery & equipment nec	4.1	8.1
Other manufacturing products	3.8	7.5

Source: GTAP

GTAP results are heavily influenced by the choice of microeconomic and macroeconomic “closure” rules[9]. Under the GTAP model’s default microeconomic closure, the factor endowments (i.e. the total supply of labour, both skilled and unskilled, as well as of capital and land) are fixed; factor prices (i.e. wages and return to capital and land) adjust to restore full employment of the factors of production in the post-shock equilibrium.[10] Under alternative microeconomic closures that are sometimes used, the return to capital or to labour can be fixed and the supply of capital and/or labour then adjusts to restore equilibrium.[11] Each of these closure rules makes an extreme assumption about the supply of labour and/or capital: it is either perfectly elastic or perfectly inelastic. The reality is likely to be somewhere in between, meaning there is likely to be some impact on factor endowments stemming from an FTA.  The economic literature supports positive but low long-run labour supply elasticity; the default closure rule is likely to be a reasonably good approximation, although the welfare impacts are likely to be somewhat understated due to missed endowment effects. With regard to the long-run supply of capital, given the high degree of globalization of capital markets, small open economies that have good access to capital are likely to face a relatively elastic capital supply, even if there is a low domestic savings response to changes in rates of return.  The closure rule under which rates of return are fixed and the supply of capital adjusts may be the better approximation, although the endowment effects are likely to be somewhat overstated.

Given these considerations, we use two alternative microeconomic closure rules: the default closure (no endowment effects) and a closure rule under which the rates of return to capital are fixed for the FTA partners and for Taiwan but the labour supply is fixed (capital flexible).  The second closure rule introduces a quasi-dynamic effect into the simulations.  The expectation is that East Asian the members of the regional FTA would experience greater positive benefits while the impact on Taiwan would be equivocal, depending on whether Taiwan’s export gains driven by the additional GDP gains in the FTA member economies outweigh any losses due to reduce levels of investment, on top of the losses due to trade diversion.

The second aspect of closure is macroeconomic closure. Two approaches are available here: the standard approach with the GTAP model, which is used in the present simulations, is to allow the current account to adjust to the trade shock, with passive accommodation by international investment flows. The change in the current account implies a change in domestic investment. In the GTAP model, the change in investment is reflected in the profile of final demand, which in turn affects the profile of production and trade but does not feed through into the productive capacity of industries/regions. The alternative macroeconomic closure is to fix the current account, implicitly assuming no international capital mobility; this is a much less realistic assumption; this option is accordingly eschewed.[12]

The simulations are subject to a number of general caveats.  CGE simulations do not take into account the impact of many elements of modern FTAs which typically address a wide range of issues over and above tariffs on merchandise trade, such as cross-border trade in services, specific financial services measures, temporary entry of businesspersons, investment, government procurement, competition, intellectual property, e-commerce, dispute settlement and institutional provisions. As well, FTAs often address non-tariff issues that affect goods trade, including non-tariff measures (such as standards), customs procedures, trade facilitation and so forth. Given complementarities between investment and services trade on the one hand and goods trade on the other, services and investment  liberalization could induce a stronger response of goods trade to an FTA than tariff elimination alone would predict. Moreover, in the context of sunk costs of market entry, the political commitment and the non-tariff facilitative aspects of an FTA can provide extra inducement to business to commit the resources to take advantage of the new market opportunities. These various considerations suggest that the estimated increase in merchandise trade generated in CGE simulations amongst FTA partners is likely to underestimate the actual increase that would take place given a high quality FTA. 

At the same time, FTAs typically include restrictive provisions for “sensitive” sectors which weaken the impact of FTAs. It is unlikely, for example, that significant agricultural sector liberalization (either with regard to tariff protection or reduction of producer support) would be included in any bilateral or regional trade agreement. Similarly, in sensitive areas of manufacturing such as textiles and apparel, rules of origin might restrict entry of products that do not meet minimum levels of domestic value-added. 

Taken together, the simulations should be taken as indicative of the order of magnitude of the potential impacts of an East Asian regional FTA on the FTA members and on those excluded and not as an attempt to provide precise predictions of the impact of such a regional FTA.

Results: Fixed Endowments Scenario

Table 2 sets out the changes in Taiwan’s merchandise exports to the world as a result of an East Asian FTA from which it is excluded. As can be seen, the simulations suggest that Taiwan’s exports would be impacted negatively by trade diversion, with particularly large impacts in the textiles (-9.5%), leather (-5.3%), chemicals (-5.3%), motor vehicles/parts (-5.1%).  At the same time, some sectors where trade protection is fairly low (e.g., electronic equipment) would benefit from the income gains in East Asia that such an FTA would generate. On balance, Taiwan’s exports would be cut by 1.1%.

Table 2: Impact of an East Asian FTA on Taiwan’s Merchandise Exports, Fixed Endowments Scenario

	Pre-FTA 2001 USD millions	Post-FTA,  2001 USD millions	Change in 2001 USD millions	% Change
Rice	0.0	0.0	0.0	0.0
Wheat	0.8	0.8	0.0	1.2
Cereal grains	0.7	0.7	0.0	1.4
Vegetables & fruit	75.3	68.4	-6.9	-9.1
Oil seeds	3.0	3.0	0.0	-1.0
Sugar	0.0	0.0	0.0	-
Plant-based fibres	7.6	7.7	0.1	1.7
Crops	145.6	133.0	-12.7	-8.7
Live animals	0.1	0.1	0.0	16.7
Animal products	144.6	146.9	2.3	1.6
Wool	21.5	23.2	1.7	7.9
Forestry	4.6	4.7	0.1	2.4
Fishing	163.6	169.7	6.1	3.7
Coal	0.0	0.0	0.0	-
Oil	0.0	0.0	0.0	0.0
Gas	0.0	0.0	0.0	-
Minerals	29.4	29.4	0.0	0.1
Bovine meat	14.3	14.4	0.2	1.1
Meat products	46.4	39.6	-6.8	-14.7
Vegetable oils	11.1	10.7	-0.4	-3.9
Dairy products	8.2	8.3	0.1	1.1
Processed rice	23.3	20.4	-2.9	-12.5
Processed Sugar	8.2	8.1	-0.1	-1.7
Food products	1,381.0	1,291.8	-89.2	-6.5
Beverages & tobacco	60.3	59.0	-1.3	-2.2
Textiles	11,426.8	10,342.5	-1,084.3	-9.5
Apparel	1,848.7	1,868.6	19.9	1.1
Leather products	1,537.2	1,456.3	-80.9	-5.3
Wood products	1,981.9	2,012.7	30.8	1.6
Paper & publishing	942.7	936.9	-5.8	-0.6
Petroleum & coal	653.4	633.6	-19.8	-3.0
Chemical products	14,253.0	13,503.3	-749.7	-5.3
Mineral products	1,591.7	1,530.9	-60.8	-3.8
Ferrous metals	2,861.4	2,776.3	-85.1	-3.0
Metals	968.4	952.5	-15.9	-1.6
Metal products	5,582.1	5,631.2	49.1	0.9
Motor vehicles & parts	2,058.2	1,952.6	-105.6	-5.1
Transport equipment	2,684.0	2,694.9	11.0	0.4
Electronic equipment	51,416.3	52,277.5	861.2	1.7
Machinery & equipment	20,095.7	19,976.6	-119.1	-0.6
Other mfg products	3,514.4	3,563.4	49.1	1.4
Total	125,565.0	124,149.5	-1,415.5	-1.1

Source: Authors’ estimates

Table 3 provides the corresponding impacts on Taiwan’s imports from the world.  As can be seen, most sectors experience a decline in imports, reflecting the negative impact on Taiwan’s income from an East Asian FTA from which it is excluded.  In contrast to the export picture, the reductions in import levels are for the most part evenly distributed, with significant impacts only in sectors which rely on imported inputs for exports and which would be negatively impacted by trade diversion.

Table 3: Impact of an East Asian FTA on Taiwan’s Merchandise Imports, Fixed Endowments Scenario

	Pre-FTA, 2001 USD millions	Post-FTA, 2001 USD millions	Change in 2001 USD millions	% Change
Rice	0.7	0.7	0.0	-5.6
Wheat	175.4	170.9	-4.5	-2.6
Cereal grains	648.2	637.0	-11.2	-1.7
Vegetables & fruit	505.4	495.4	-10.1	-2.0
Oil seeds	506.7	502.0	-4.7	-0.9
Sugar	0.3	0.3	0.0	3.2
Plant-based fibres	321.1	296.0	-25.1	-7.8
Crops	358.2	349.3	-8.9	-2.5
Live animals	1.3	1.3	0.0	-0.8
Animal products	399.4	385.1	-14.3	-3.6
Wool	92.1	85.2	-6.8	-7.4
Forestry	190.4	193.6	3.2	1.7
Fishing	80.8	79.9	-0.9	-1.2
Coal	2,041.1	2,038.1	-2.9	-0.1
Oil	5,113.6	5,056.6	-57.0	-1.1
Gas	648.4	650.5	2.1	0.3
Minerals	758.9	755.1	-3.8	-0.5
Bovine meat	249.3	244.0	-5.2	-2.1
Meat products	84.8	80.6	-4.2	-5.0
Vegetable oils	87.2	84.3	-2.9	-3.4
Dairy products	280.2	275.0	-5.2	-1.9
Processed rice	7.2	16.2	9.0	124.2
Processed Sugar	110.0	106.1	-3.9	-3.6
Food products	1,480.9	1,461.9	-19.0	-1.3
Beverages & tobacco	797.6	789.8	-7.7	-1.0
Textiles	1,440.4	1,346.9	-93.5	-6.5
Apparel	800.3	782.3	-18.0	-2.2
Leather products	539.7	522.2	-17.5	-3.2
Wood products	1,055.8	1,063.8	8.0	0.8
Paper & publishing	1,838.0	1,806.0	-32.0	-1.7
Petroleum & coal	1,493.3	1,474.9	-18.4	-1.2
Chemical products	13,111.0	12,796.5	-314.5	-2.4
Mineral products	2,182.4	2,174.7	-7.8	-0.4
Ferrous metals	3,625.0	3,588.9	-36.0	-1.0
Metals	4,043.3	4,052.0	8.7	0.2
Metal products	1,252.8	1,230.0	-22.8	-1.8
Motor vehicles & parts	2,414.3	2,371.2	-43.1	-1.8
Transport equipment	2,965.4	2,936.7	-28.7	-1.0
Electronic equipment	28,379.0	28,653.1	274.1	1.0
Machinery & equipment	22,762.3	22,489.6	-272.7	-1.2
Other mfg products	1,296.3	1,274.3	-22.0	-1.7
Total	104,138.1	103,317.6	-820.5	-0.8

Source: Authors’ estimates

As can be seen in Table 4 below, the simulations suggest that Taiwan would not be able to make up for the loss of export markets in East Asia through expanded exports to the rest of the world. This reflects the fact that the trade and income gains from an East Asian FTA would be concentrated in the members to the FTA while third parties generally experience some reduction in both trade and income.

Table 4: Changes in the Direction of Taiwan’s Trade, 2001 USD millions

Change in Taiwan's Exports
East Asian FTA Members	-3,178
Rest of the World	1,762
Total	-1,415
Change in Taiwan's Imports
East Asian FTA Members	-1,235
Rest of the world	415
Total	-821

Source: Authors’ estimates

Table 5 compares the changes in GDP on the various regions in the global economy as a result of an East Asian FTA. As can be seen, the East Asian FTA members experience a significant gain in income while Taiwan experiences a 1% reduction.  Elsewhere, the effects are negative with the largest losses in dollar terms being in the largest economies,

the United States and the 27-member European Union, and the largest percentage declines being in India and the Closer Economic Cooperation partners, Australia and New Zealand.

Table 5: Impacts on the value of GDP, Selected Economies

Region	Impact on GDP US$ 2001	Impact on GDP, %
Taiwan	-2,900.7	-1.0
East Asian FTA Members
Japan	9,695.5	0.2
China	15,881.5	1.5
Hong Kong	1,622.6	1.0
South Korea	1,404.4	0.3
ASEAN6	7,151.1	1.3
Other Economies
USA	-32,326.0	-0.3
Canada	-1,468.6	-0.2
Mexico	-920.7	-0.2
EU27	-19,742.0	-0.2
EFTA	-973.8	-0.2
Australia and New Zealand	-2,572.8	-0.6
India	-2,163.7	-0.5
Turkey	-312.1	-0.2
Rest of the Middle East	-1,713.9	-0.3
Egypt & Maghreb	-609.2	-0.3
South African Customs Union	-403.1	-0.3
Russian Federation	-1,335.2	-0.4
Rest of the Former USSR	-369.3	-0.4
Rest of Europe	-61.0	-0.1
MERCOSUR	-3,144.2	-0.3
Andean Community	-477.2	-0.3
Chile	-304.9	-0.5
Central America	-263.6	-0.4
CARICOM	-310.3	-0.3
Rest of the World	-2,079.7	-0.4
Total	-62,193.6	-0.2

Source: Authors’ estimates

The GDP impacts reported above are, as is usual with GTAP simulations, calculated on the basis of post-FTA prices and thus reflect the change in the value of GDP.  Measured at pre-FTA prices, the change in the level of economic activity in Taiwan is much smaller, only $153 million (in 2001US dollars).

Results: Capital flexible scenario

The major impacts of alternative microeconomic closure rules are on the level of income gains and the extent of trade diversion; bilateral trade figures tend to be impacted only modestly.  This section reports the results under the closure rule which fixes the rate of return to capital in the economies that are party to the FTA and in Taiwan and allows the level of capital to adjust to restore equilibrium. This scenario introduces some element of a dynamic investment response to the changed economic situation engendered by the East Asian FTA.

Table 6 sets out the changes in Taiwan’s merchandise exports to the world as a result of an East Asian FTA from which it is excluded, under the flexible capital scenario. Generally, the impacts on Taiwan’s exports are more negative when investment is allowed to respond to the FTA; with a few exceptions, those sectors that lose, lose more and those sectors that gain, gain less. Overall, the simulations suggest that Taiwan’s exports would be reduced by 1.44% compared to 1.11% when investment is not allowed to respond. 

Table 6: Impact of an East Asian FTA on Taiwan’s Merchandise Exports, Flexible Capital Scenario

(i)	(ii)       Pre-FTA (iii)      2001 USD millions	(iv)      Post-FTA, (v)         2001 USD millions	(vi)      Change in (vii)    2001 USD millions	(viii)   % Change
(ix)      Rice	(x)        0.0	(xi)      0.0	(xii)    0.0	(xiii)   0.0
(xiv)   Wheat	(xv)     0.8	(xvi)   0.8	(xvii) 0.0	(xviii)              2.5
(xix)   Cereal grains	(xx)     0.7	(xxi)   0.7	(xxii) 0.0	(xxiii)              2.8
(xxiv)              Vegetables & fruit	(xxv)  75.3	(xxvi)              68.9	(xxvii)             -6.4	(xxviii)           -8.5
(xxix)              Oil seeds	(xxx)  3.0	(xxxi)              3.0	(xxxii)             0.0	(xxxiii)           0.7
(xxxiv)           Sugar	(xxxv)             0.0	(xxxvi)           0.0	(xxxvii)          0.0	(xxxviii)        0.0
(xxxix)           Plant-based fibres	(xl)      7.6	(xli)    8.0	(xlii)   0.4	(xliii) 5.4
(xliv) Crops	(xlv)   145.6	(xlvi) 135.7	(xlvii)              -9.9	(xlviii)            -6.8
(xlix) Live animals	(l)         0.1	(li)       0.1	(lii)      0.0	(liii)    16.7
(liv)    Animal products	(lv)      144.6	(lvi)    147.8	(lvii)   3.2	(lviii) 2.2
(lix)    Wool	(lx)      21.5	(lxi)    23.6	(lxii)   2.1	(lxiii) 9.8
(lxiv) Forestry	(lxv)   4.6	(lxvi) 4.8	(lxvii)              0.2	(lxviii)            3.9
(lxix) Fishing	(lxx)   163.6	(lxxi) 173.2	(lxxii)              9.6	(lxxiii)            5.9
(lxxiv)             Coal	(lxxv)              0.0	(lxxvi)             0.0	(lxxvii)           0.0	(lxxviii)         0.0
(lxxix)             Oil	(lxxx)              0.0	(lxxxi)             0.0	(lxxxii)           0.0	(lxxxiii)         0.0
(lxxxiv)          Gas	(lxxxv)           0.0	(lxxxvi)          0.0	(lxxxvii)        0.0	(lxxxviii)      0.0
(lxxxix)          Minerals	(xc)     29.4	(xci)   29.9	(xcii)  0.5	(xciii)              1.7
(xciv)               Bovine meat	(xcv)  14.3	(xcvi)               14.5	(xcvii)             0.3	(xcviii)           2.0
(xcix)               Meat products	(c)        46.4	(ci)      39.7	(cii)     -6.7	(ciii)   -14.4
(civ)   Vegetable oils	(cv)     11.1	(cvi)   10.6	(cvii)  -0.6	(cviii)              -5.0
(cix)   Dairy products	(cx)     8.2	(cxi)   8.3	(cxii)  0.2	(cxiii)              2.0
(cxiv)               Processed rice	(cxv)  23.3	(cxvi)               20.6	(cxvii)             -2.7	(cxviii)           -11.5
(cxix)               Processed Sugar	(cxx)  8.2	(cxxi)               8.1	(cxxii)             -0.1	(cxxiii)           -0.9
(cxxiv)            Food products	(cxxv)             1,381.0	(cxxvi)            1,292.4	(cxxvii)          -88.6	(cxxviii)        -6.4
(cxxix)            Beverages & tobacco	(cxxx)             60.3	(cxxxi)            59.1	(cxxxii)          -1.2	(cxxxiii)        -2.0
(cxxxiv)         Textiles	(cxxxv)          11,426.8	(cxxxvi)         10,196.8	(cxxxvii)       -1,230.0	(cxxxviii)     -10.8
(cxxxix)         Apparel	(cxl)   1,848.7	(cxli)  1,873.2	(cxlii)              24.5	(cxliii)             1.3
(cxliv)             Leather products	(cxlv)               1,537.2	(cxlvi)             1,453.1	(cxlvii)           -84.1	(cxlviii)          -5.5
(cxlix)             Wood products	(cl)      1,981.9	(cli)     2,022.7	(clii)   40.9	(cliii) 2.1
(cliv)  Paper & publishing	(clv)   942.7	(clvi)  934.8	(clvii)              -7.8	(clviii)             -0.8
(clix)  Petroleum & coal	(clx)   653.4	(clxi)  639.0	(clxii)              -14.3	(clxiii)             -2.2
(clxiv)             Chemical products	(clxv)               14,253.0	(clxvi)             13,410.0	(clxvii)           -843.0	(clxviii)          -5.9
(clxix)             Mineral products	(clxx)               1,591.7	(clxxi)             1,531.8	(clxxii)           -59.8	(clxxiii)          -3.8
(clxxiv)          Ferrous metals	(clxxv)            2,861.4	(clxxvi)          2,750.6	(clxxvii)        -110.7	(clxxviii)       -3.9
(clxxix)          Metals	(clxxx)            968.4	(clxxxi)          957.9	(clxxxii)        -10.5	(clxxxiii)       -1.1
(clxxxiv)       Metal products	(clxxxv)         5,582.1	(clxxxvi)       5,640.5	(clxxxvii)     58.4	(clxxxviii)    1.0
(clxxxix)       Motor vehicles & parts	(cxc)  2,058.2	(cxci) 1,947.5	(cxcii)             -110.6	(cxciii)            -5.4
(cxciv)            Transport equipment	(cxcv)              2,684.0	(cxcvi)            2,702.9	(cxcvii)          19.0	(cxcviii)         0.7
(cxcix)            Electronic equipment	(cc)     51,416.3	(cci)    52,077.7	(ccii)  661.4	(cciii)               1.3
(cciv) Machinery & equipment	(ccv)  20,095.7	(ccvi) 19,975.5	(ccvii)             -120.3	(ccviii)            -0.6
(ccix) Other mfg products	(ccx)  3,514.4	(ccxi) 3,591.2	(ccxii)             76.8	(ccxiii)            2.2
(ccxiv)            Total	(ccxv)              125,565.0	(ccxvi)            123,755.1	(ccxvii)          -1,809.9	(ccxviii)         -1.4

Source: Authors’ estimates

Table 7 provides the corresponding impacts on Taiwan’s imports from the world under the flexible capital scenario.  The overall impact on imports is similar to that in the fixed endowments scenario; the sectoral composition is also little changed.

Table 7: Impact of an East Asian FTA on Taiwan’s Merchandise Imports, Flexible Capital Scenario

(ccxix)	(ccxx)              Pre-FTA, (ccxxi)            2001 USD millions	(ccxxii)          Post-FTA, (ccxxiii)         2001 USD millions	(ccxxiv)         Change in (ccxxv)           2001 USD millions	(ccxxvi)         % Change
(ccxxvii)       Rice	(ccxxviii)      0.7	(ccxxix)         0.7	(ccxxx)           0.0	(ccxxxi)         -7.0
(ccxxxii)       Wheat	(ccxxxiii)      175.4	(ccxxxiv)      171.2	(ccxxxv)        -4.2	(ccxxxvi)      -2.4
(ccxxxvii)    Cereal grains	(ccxxxviii)   648.2	(ccxxxix)      637.5	(ccxl) -10.7	(ccxli)             -1.7
(ccxlii)            Vegetables & fruit	(ccxliii)          505.4	(ccxliv)          494.8	(ccxlv)            -10.6	(ccxlvi)          -2.1
(ccxlvii)         Oil seeds	(ccxlviii)       506.7	(ccxlix)          502.5	(ccl)    -4.3	(ccli)  -0.8
(cclii)               Sugar	(ccliii)             0.3	(ccliv)             0.3	(cclv) 0.0	(cclvi)             3.2
(cclvii)            Plant-based fibres	(cclviii)          321.1	(cclix)             292.6	(cclx) -28.5	(cclxi)             -8.9
(cclxii)            Crops	(cclxiii)          358.2	(cclxiv)          348.5	(cclxv)            -9.7	(cclxvi)          -2.7
(cclxvii)         Live animals	(cclxviii)       1.3	(cclxix)          1.3	(cclxx)            0.0	(cclxxi)          -0.8
(cclxxii)         Animal products	(cclxxiii)       399.4	(cclxxiv)       384.6	(cclxxv)         -14.8	(cclxxvi)       -3.7
(cclxxvii)      Wool	(cclxxviii)    92.1	(cclxxix)       84.2	(cclxxx)         -7.8	(cclxxxi)       -8.5
(cclxxxii)      Forestry	(cclxxxiii)    190.4	(cclxxxiv)    194.0	(cclxxxv)      3.6	(cclxxxvi)    1.9
(cclxxxvii)   Fishing	(cclxxxviii) 80.8	(cclxxxix)    79.9	(ccxc)              -0.9	(ccxci)            -1.1
(ccxcii)           Coal	(ccxciii)         2,041.1	(ccxciv)         2,047.7	(ccxcv)           6.7	(ccxcvi)         0.3
(ccxcvii)        Oil	(ccxcviii)      5,113.6	(ccxcix)         5,080.1	(ccc)   -33.5	(ccci) -0.7
(cccii)              Gas	(ccciii)            648.4	(ccciv)            653.8	(cccv)              5.4	(cccvi)            0.8
(cccvii)           Minerals	(cccviii)         758.9	(cccix)            753.4	(cccx)              -5.5	(cccxi)            -0.7
(cccxii)           Bovine meat	(cccxiii)         249.3	(cccxiv)         243.1	(cccxv)           -6.1	(cccxvi)         -2.5
(cccxvii)        Meat products	(cccxviii)      84.8	(cccxix)         80.5	(cccxx)           -4.4	(cccxxi)         -5.1
(cccxxii)        Vegetable oils	(cccxxiii)      87.2	(cccxxiv)      85.0	(cccxxv)        -2.2	(cccxxvi)      -2.6
(cccxxvii)     Dairy products	(cccxxviii)   280.2	(cccxxix)      274.3	(cccxxx)        -5.9	(cccxxxi)      -2.1
(cccxxxii)     Processed rice	(cccxxxiii)   7.2	(cccxxxiv)   16.3	(cccxxxv)     9.1	(cccxxxvi)   125.3
(cccxxxvii)  Processed Sugar	(cccxxxviii)              110.0	(cccxxxix)   106.0	(cccxl)            -4.0	(cccxli)           -3.6
(cccxlii)         Food products	(cccxliii)       1,480.9	(cccxliv)        1,462.2	(cccxlv)         -18.7	(cccxlvi)        -1.3
(cccxlvii)      Beverages & tobacco	(cccxlviii)    797.6	(cccxlix)        789.5	(cccl) -8.1	(cccli)              -1.0
(ccclii)            Textiles	(cccliii)          1,440.4	(cccliv)           1,345.3	(ccclv)            -95.1	(ccclvi)           -6.6
(ccclvii)         Apparel	(ccclviii)       800.3	(ccclix)           783.5	(ccclx)            -16.8	(ccclxi)           -2.1
(ccclxii)         Leather products	(ccclxiii)       539.7	(ccclxiv)        521.7	(ccclxv)         -18.0	(ccclxvi)        -3.3
(ccclxvii)      Wood products	(ccclxviii)    1,055.8	(ccclxix)        1,065.3	(ccclxx)         9.4	(ccclxxi)        0.9
(ccclxxii)      Paper & publishing	(ccclxxiii)    1,838.0	(ccclxxiv)     1,805.5	(ccclxxv)      -32.5	(ccclxxvi)     -1.8
(ccclxxvii)   Petroleum & coal	(ccclxxviii) 1,493.3	(ccclxxix)     1,480.8	(ccclxxx)      -12.5	(ccclxxxi)     -0.8
(ccclxxxii)   Chemical products	(ccclxxxiii) 13,111.0	(ccclxxxiv)  12,777.9	(ccclxxxv)   -333.0	(ccclxxxvi)  -2.5
(ccclxxxvii)              Mineral products	(ccclxxxviii)             2,182.4	(ccclxxxix)  2,174.3	(cccxc)           -8.1	(cccxci)          -0.4
(cccxcii)        Ferrous metals	(cccxciii)      3,625.0	(cccxciv)       3,588.4	(cccxcv)        -36.5	(cccxcvi)       -1.0
(cccxcvii)     Metals	(cccxcviii)   4,043.3	(cccxcix)       4,043.6	(cd)     0.3	(cdi)   0.0
(cdii)  Metal products	(cdiii)              1,252.8	(cdiv)               1,234.2	(cdv)  -18.5	(cdvi)               -1.5
(cdvii)             Motor vehicles & parts	(cdviii)           2,414.3	(cdix)               2,375.3	(cdx)  -38.9	(cdxi)               -1.6
(cdxii)             Transport equipment	(cdxiii)           2,965.4	(cdxiv)            2,943.7	(cdxv)             -21.7	(cdxvi)            -0.7
(cdxvii)          Electronic equipment	(cdxviii)        28,379.0	(cdxix)            28,571.2	(cdxx)             192.2	(cdxxi)            0.7
(cdxxii)          Machinery & equipment	(cdxxiii)        22,762.3	(cdxxiv)         22,545.4	(cdxxv)          -216.9	(cdxxvi)         -1.0
(cdxxvii)       Other mfg products	(cdxxviii)     1,296.3	(cdxxix)         1,275.2	(cdxxx)          -21.1	(cdxxxi)         -1.6
(cdxxxii)       Total	(cdxxxiii)     104,138.1	(cdxxxiv)      103,315.1	(cdxxxv)       -823.0	(cdxxxvi)      -0.8

Source: Authors’ estimates

As can be seen in Table 8 below, introducing some element of dynamism does not change significantly the overall impact of an East Asian FTA on Taiwan. The simulations continue to suggest that Taiwan would not be able to make up for the loss of export markets in East Asia through expanded exports to the rest of the world notwithstanding the fact that the negative impact on the global economy is reduced when dynamic responses are taken into account.

Table 8: Changes in the Direction of Taiwan’s Trade, 2001 USD millions,

Flexible Capital Scenario

Change in Taiwan's Exports
East Asian FTA Members	-3,753
Rest of the World	1,943
Total	-1,810
Change in Taiwan's Imports
East Asian FTA Members	-159
Rest of the world	-664
Total	-823

Source: Authors’ estimates

Table 9 compares the changes in GDP in the various regions of the global economy as a result of an East Asian FTA. In sharp contrast to the fixed endowments scenario, the global economy is not negatively impacted by the East Asian FTA. However, while the negative impact worldwide is diminished, and the GDP gains in the FTA member economies is greatly strengthened (especially in South Korea, China and the ASEAN economies), the negative impact on Taiwan is marginally exacerbated, with the GDP loss rising to -1.14% from -1.0% in the fixed endowments scenario.

Table 9: Impacts on the value of GDP, Selected Economies,

Flexible Capital Scenario

(cdxxxvii)    Region	(cdxxxviii)  Impact on GDP US$ 2001	(cdxxxix)      Impact on GDP, %
(cdxl)              Taiwan	(cdxli)             -3,246	(cdxlii)           -1.14
(cdxliii)          East Asian FTA Members	(cdxliv)	(cdxlv)
(cdxlvi)            Japan	(cdxlvii)        10,344	(cdxlviii)       0.25
(cdxlix)            China	(cdl)   9,212	(cdli)  0.84
(cdlii)  Hong Kong	(cdliii)             2,398	(cdliv)             1.48
(cdlv)   South Korea	(cdlvi)             23,738	(cdlvii)           5.54
(cdlviii)            ASEAN6	(cdlix)             16,509	(cdlx)               3.08
(cdlxi)             Other Economies	(cdlxii)	(cdlxiii)
(cdlxiv)            USA	(cdlxv)            -14,670	(cdlxvi)          -0.15
(cdlxvii)           Canada	(cdlxviii)       -311	(cdlxix)          -0.04
(cdlxx)              Mexico	(cdlxxi)          -292	(cdlxxii)        -0.05
(cdlxxiii)         EU27	(cdlxxiv)       -7,146	(cdlxxv)         -0.09
(cdlxxvi)         EFTA	(cdlxxvii)     -33	(cdlxxviii)    -0.01
(cdlxxix)         Australia and New Zealand	(cdlxxx)         -1,081	(cdlxxxi)       -0.26
(cdlxxxii)        India	(cdlxxxiii)    -1,334	(cdlxxxiv)    -0.28
(cdlxxxv)        Turkey	(cdlxxxvi)    -119	(cdlxxxvii)  -0.08
(cdlxxxviii)   Rest of the Middle East	(cdlxxxix)    2,180	(cdxc)              0.34
(cdxci)              Egypt & Maghreb	(cdxcii)          375	(cdxciii)         0.16
(cdxciv)           South African Customs Union	(cdxcv)           -130	(cdxcvi)         -0.11
(cdxcvii)          Russian Federation	(cdxcviii)      -197	(cdxcix)         -0.06
(d)          Rest of the Former USSR	(di)      -155	(dii)    -0.15
(diii)     Rest of Europe	(div)   8	(dv)     0.02
(dvi)     MERCOSUR	(dvii) 315	(dviii)              0.03
(dix)     Andean Community	(dx)     72	(dxi)   0.04
(dxii)    Chile	(dxiii)              -144	(dxiv)              -0.22
(dxv)    Central America	(dxvi)              -40	(dxvii)             -0.06
(dxviii)             CARICOM	(dxix)              -19	(dxx)  -0.02
(dxxi)  Rest of the World	(dxxii)             -416	(dxxiii)           -0.08
(dxxiv)           Total	(dxxv)             35,818	(dxxvi)           0.11

Source: Authors’ estimates

Comparisons with previous studies

One recent study by Chen and Ku (2007) showed similar results as what is found in this study. Though the classification of product groups in this study is different from that of Chen and Ku ( 2007), one can still identify some interesting comparison ; the negative effect on Taiwan’s GDP is more significant under “dynamic model” than that under “ static model” in either ASEAN plus 3 ( AP3) or ASEAN plus China ( AP1). ). In the dynamic model simulations by Chen and Ku ( 2007), Taiwan would suffer from a loss of its GDP by -0.114% under ASEAN plus China ( AP1) , and -0.473% under ASEAN plus 3 ( AP3).[13]  The report on Table 9 showed that Taiwan’s GDP will suffer from a negative 1.14% at 2001 constant US dollar.

At sectoral levels, Chen and Ku (2007) found that the adversary effects of ASEAN plus 3 (AP3) on industry outputs of Taiwan investigated are most significant in the textile sector (-8.89%), food (-1.92%), plastics (-1.67%), apparel (-1.44%). Taiwan’s automobile outputs will be negatively affected by ASEAN plus 3 by a -1.05%, but outputs of other transport equipment will increase by 1.42%. Their finding that “electronic equipment” output has a positive growth of 2.64% is also consistent the 1.7% of export growth found in Table 2 of this study.

Henceforth, this finding, even considering the new database and new development of the CEPA between China and Hong Kong, is fairly consistent with Chen and Ku (2007)

III. Export Potential of Taiwan in the US and ASEAN Plus 6 from TradeSim Gravity Model Simulations

The third version of TradeSim updated the gravity model by using the 2002 and 2003 trade data to estimate trade potentials in 19 sectors for 132 exporting and 154 importing countries. The data source is from COMTRADE, the UN trade database.[14] For the bilateral measure of market access, one can find from Bouet et al (2001 about the construction and the interpretation of the bilateral database, called the “MacMap”...[15] The model specification was suggested by Anderson and van Wincoop (2003), and the following equation was estimated by using a Poisson Pseudo Maximum Likelihood (PPMLE) procedure proposed by Santos Silva and Tenreyro (2004): [16]

X =αi k: +αj k +β0k +β1k D +β2k Tariff +β3k Border i j  +β4k Language i j  +β 5kConflict i j +β6k Geo i j +ε

where:

i  : the exporting country

j: the importing country

k: sector

Xijk : trade in sector k from country i to country j

Dij: distance between i and j

Borderij: i and j are neighboring countries (=1) or not (=0)

Tariffijk: bilateral market access measure (for trade from i to j in sector k)

Languageij: bilateral measure of common language

Conflictij: bilateral measure of conflict

ij Geo : bilateral measure of geographical location

ik jk α ;α : multilateral resistance terms in form of fixed effects.

Export potentials are interpreted as “potentials within the specified model set-up”;[17] 

They represent the “relative trade potentials”, which are “independent of the absolute value of exports”. Hence, “highly untapped potential does not imply that this potential is high in absolute terms but that is high compared to the current export”. In spite of the caveats, [18] simulation results from TradeSim, after complemented with a deeper sector-specific analysis, could capture the trade complementarities between countries.

Export potentials for Taiwan were estimated for 15 sectors and 30 major trading partners in the world.[19] Relative trade potential in each sector is ranked into 5 classifications in an ascending order of trading opportunities: 1 stands for “very strong current trade that is above predicted value. 2 stands for strong current trade, that is also above the predicted. 3 stands for predicted values equal to current or low ones. 4 stands for untapped trade potentials and 5 stand for high untapped trade potentials.

 Table 10 reported those sectors which have either the high untapped trade potential (5) or untapped trade potentials (4) for Taiwan’s exports to the U.S. and ASEAN plus 6. Given the qualifications of gravity model, one can identify Taiwan’s export potential in those major trading partners in the region.

{Table 10 here}

IV. Niche Products in Niche Markets: Taiwan’s Potential Trade Growth amid East Asian Economic Integration

The adversary effects of  ASEAN6 plus 3 as reported in Table 2 indicate that Taiwan will suffer from being excluded from Asian economic integration at the textile ( -9.5%), food products( -6.5%), leather products ( -5.3%), chemical products ( -5.3%), motor vehicles and parts (-5.1%), petroleum & coal  ( -3.0%), ferrous products (-3.0%), and others. However, if Taiwan could identify its niche products in niche markets, then Taiwan could still exploit its export potentials by penetrating in the world market with appropriate trade strategies.

Taiwan’s Export potential

One can find from Table 10 that in terms of trading partners, Taiwan has highly export potentials in Korea and moderate untapped trade potentials with Japan, Malaysia and the Philippines. For Hong Kong, Indonesia, Singapore, Thailand and the U.S., the predicted values from gravity model are equal to current or low levels.  For China and Vietnam, Taiwan has strong current trade above the predicted values. The simulation results may challenge the intuition a little bit, especially on Taiwan’s export potential to China and Vietnam. Common sense suggests that “FDI induced export” has accounted for a significant portion of Taiwan’s export to China, Vietnam and other hosting countries. TradeSim model has two sets of sample countries where FDI data are available at aggregate (29 exporting and 165 importing countries) or disaggregate levels (7 exporting and 65 importing countries). Since the model of simulations for Taiwan’s export potentials does not belong to either set, the simulation results reported here may underestimate Taiwan’s trade potential by ignoring the contribution of its outward FDI on export potentials.

On sectoral and country-specific basis, Taiwan has 2 sectors which have untapped trade potential in the U.S; chemicals/chemical products, which has a share of 8.3% in Taiwan’s exports of the sector, and coke / petroleum products / nuclear fuels, which accounted for 4.5% of Taiwan’s export in that sector. Meanwhile, the tariffs applied by the U.S. in these two sectors are very low from zero to 5%. It implies that Taiwan has the opportunity to exploit its export in these two sectors the U.S. market. In its trade with Hong Kong, Taiwan has high untapped trade potentials in “other manufacturing”, which accounted for 8% in Taiwan’s exports in that sector. As a free entreport, Hong Kong also has a low tariffs structure from zero to 5% too. Though Taiwan has “untapped export potential” in Malaysia, its export potential is concentrated in “electric and electronic equipment” sector, which also has low tariffs from zero to 5% in the importing country...

A similar situation occurred in Taiwan trade potential with the Philippines; Taiwan has untapped trade potential with that country, yet its export potentials are concentrated in  “ electric and electronic equipment” which accounts for 1.3% of Taiwan export in the sector, and in “ metal and metal products” sector which accounted for 1 % in Taiwan’s export in that sector. The Philippines have low tariffs in these two sectors from zero to 5% in these two sectors. It implies that there are substantial opportunities for Taiwan to exploit its trade potential in these sectors which have low tariffs in these importing countries even after the regional trading blocs excludes Taiwan from it.

In Japanese market, Taiwan has untapped export potentials in 8 sectors; which have low tariffs from zero to 5%, except for textile/clothing/ leather sector which has 20% to 25%. Taiwan has “high untapped trade potentials in “chemical/ chemical products”, “precision instruments” and “textile/clothing/ leather products”.  The first two sectors accounted for 6% and 8% of Taiwan’s export in these sectors respectively. For  “electric/electronic products”, “ machinery and equipment”, “rubber and plastic products”, motor vehicles and other transport equipment”,“ wood/wood products” and “ coke/petroleum products/ nuclear fuel”, Taiwan has “ untapped trade potential” , which account for significant percentage shares in Taiwan exports, and have low tariffs from zero to 5% in Japanese market. Again, Taiwan has yet to fully exploit its trade potential in Japan.  

For trade with Korea, Taiwan has 3 sectors which it has “ high untapped trade sectors” in “ chemical/ chemical products”, “ metal/ metal products” and machinery and equipment”. All three sectors have tariffs from 5% to 10% in Korea. For “electric/electronic products”,” precision instruments”, “ textile/clothing/ leather products”, “rubber and plastic products”,” food/beverage/tobacco products”, “motor vehicles / other transport equipment”, and “ wood/wood products, Taiwan also has “ untapped trade potentials” in Korea. Except for” food/beverage/tobacco” and “textile/clothing/ leather” sectors , which have 35 - 40% and 10-15% of tariffs respectively, all of them have only a maximum of 5 of tariffs in Korea. In Singapore, Taiwan has 4 sectors which have “untapped trade potentials” in “chemical/ chemical products”,” metal/ metal products”, “machinery and equipment”, and “motor vehicles / other transport equipment”. All of them have a zero to 5% of tariffs in Singapore.

Therefore, sector by sector, one can find from Table 6 that Taiwan has untapped trade potentials on electric and electronic products in Japan (which accounted for 9.5% of its total exports in that sector), Korea (4.2%), Malaysia (2.9%), the Philippines (1.3%) and Thailand (1.3%). These 5 importing countries accounted for a total of 19.4% of Taiwan’s total exports in that sector, which is a sizable portion and formidable opportunity for Taiwan’s export business in that sector. Another sector is Taiwan’s export of chemical products; If Taiwan could penetrate into the U.S. (8.3%). Japan (5.9%), Korea (2.6%) and Singapore (1.4%), then altogether 18.2% of Taiwan’s total exports in chemical products still have the untapped trade potentials in these 4 importing countries. Other sectors which Taiwan has untapped trade potentials are “precision instruments “ ( 12.8%) in Japan ( 8%), Korea ( 3.6%) and Singapore ( 1.2%) ; wood products( 10.8%)  in Japan ( 10.8%) exclusively, motor vehicle and other transport equipment ( 10.5%) in Japan ( 7%), Singapore ( 3%) and Korea ( 1.5%);  Coke et al ( 104%) in the U.S. ( 4.5%) and Japan ( 5.9%); rubber and plastic products ( 9.7%) in Japan ( 8.7%) and Korea ( 1%); machinery ( 8.5%) in Japan ( 5.8%), Singapore ( 1.4%) and Korea ( 1.3%); other manufacturing products ( 8%) in Hong Kong ( 8%) exclusively.

Even in textile sector which Taiwan will suffer the most from East economic integration (-9.5%), Taiwan still has untapped export potential (3.8%) in Japan (2.8%) and Korea (1%).

V. Summary and Conclusions

This study investigates Taiwan’s risks posed by exclusion from East Asia’s emerging set of FTAs and the opportunities from trade potential by trade partner and by commodity sectors. Even in the probably worst scenarios of a comprehensive FTA involving the ASEAN plus 3, Taiwan could still capture its export potential by penetrating in those “niche markets” in those “niche products”. The emerging East Asian economic integration posed some critical challenges to Taiwan future export growth, yet model simulations from TradeSim show that there are tremendous opportunities for Taiwan to exploit its export potentials by pursuing an appropriate trade strategy by identifying its competitiveness in both the “niche products” in the “niche markets”.

Given the reality of international politics, Taiwan may not be able to sign many FTAS with its trading partners as she wishes, yet it does not preclude Taiwan to exploit its trade potential potentials, which if appropriately implemented, will be able to mitigate the adversary effect of being excluded from those trading blocs.

The TradeSim model only investigates the Taiwan’s trade potentials based on the gravity model without including complementary effects of outward FDI on trade flows. It definitely underestimates Taiwan’s trade potential because trade flows induced by the FDI was excluded from the model simulations. As Taiwan has become one of the largest investors in the world economy, further studies on trade potentials need to consider the complementary effects of outward FDI on trade flows.  

Table 10: Current Export and Export Potential of Taiwan to the US and ASEAN plus 6 (US million $ unless specified).

Current exports and export potential of Taiwan  to its main markets (USD million unless specified)
Exporter	Rank	Importer	Rank	Sector	Current Exports 2002-2003	share in TWN 's exports of sector, in %	Relative Trade Potential	Total FDI outward stock 2003	Share in Total FDI Outstock, in %	Tariff applied by importer, in %
	1	United States		TOTAL	28,262,045	19.40%	3. Predicted = Current or low values	.	.
Taiwan	1	United States	8	Chemicals and chemical products	1,223,917	8.30%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	1	United States	14	Coke, petroleum products and nuclear fuel	77,178	4.50%	4. Untapped trade potential	.	.	0 to 5%
	2	China		TOTAL	25,018,313	17.20%	2. Strong current trade (above predicted)	.	.
Taiwan	2	China	14	Food, beverages and tobacco	24,999	1.60%	4. Untapped trade potential	.	.	20 to 25%
	3	Hong Kong, China		TOTAL	24,888,706	17.10%	3. Predicted = Current or low values	.	.
Taiwan	3	Hong Kong, China	8	Other manufacturing	325,860	8.00%	5. High untapped trade potential	.	.	0 to 5%
	4	Japan		TOTAL	12,360,817	8.50%	4. Untapped trade potential	.	.
Taiwan	4	Japan	1	Electrical and electronic equipment	6,378,695	9.50%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	2	Chemicals and chemical products	871,075	5.90%	5. High untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	5	Machinery and equipment	649,641	5.80%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	7	Precision instruments	510,791	8.00%	5. High untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	8	Rubber and plastic products	404,645	8.70%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	9	Motor vehicles and other transport equipment	350,563	7.00%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	10	Textiles, clothing and leather	323,353	2.80%	5. High untapped trade potential	.	.	20 to 25%
Taiwan	4	Japan	12	Wood and wood products	136,583	10.80%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	4	Japan	15	Coke, petroleum products and nuclear fuel	100,620	5.90%	4. Untapped trade potential	.	.	> 50%
	5	Singapore		TOTAL	5,068,045	3.50%	3. Predicted = Current or low values
Taiwan	5	Singapore	2	Chemicals and chemical products	203,768	1.40%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	5	Singapore	3	Metal and metal products	201,249	1.50%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	5	Singapore	5	Machinery and equipment	154,746	1.40%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	5	Singapore	7	Precision instruments	76,017	1.20%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	5	Singapore	8	Motor vehicles and other transport equipment	74,859	1.50%	4. Untapped trade potential	.	.	0 to 5%
	6	Republic of Korea		TOTAL	4,600,996	3.20%	5. High untapped trade potential
Taiwan	6	Republic of Korea	1	Electrical and electronic equipment	2,795,445	4.20%	4. Untapped trade potential	.	.	0 to 5%
Taiwan	6	Republic of Korea	3	Chemicals and chemical products	381,342	2.60%	5. High untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	4	Metal and metal products	233,675	1.70%	5. High untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	5	Precision instruments	231,537	3.60%	4. Untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	6	Machinery and equipment	149,230	1.30%	5. High untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	7	Textiles, clothing and leather	121,166	1.00%	4. Untapped trade potential	.	.	10 to 15%
Taiwan	6	Republic of Korea	9	Rubber and plastic products	44,158	1.00%	4. Untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	10	Food, beverages and tobacco	38,238	2.40%	4. Untapped trade potential	.	.	35 to 40%
Taiwan	6	Republic of Korea	11	Motor vehicles and other transport equipment	23,181	0.50%	4. Untapped trade potential	.	.	5 to 10%
Taiwan	6	Republic of Korea	14	Wood and wood products	18,667	1.50%	4. Untapped trade potential	.	.	5 to 10%
	9	Malaysia		TOTAL	3,446,833	2.40%	3. Predicted = Current or low values
Taiwan	9	Malaysia	1	Electrical and electronic equipment	1,921,564	2.90%	4. Untapped trade potential	.	.	0 to 5%
	11	Thailand		TOTAL	2,690,671	1.90%	3. Predicted = Current or low values
Taiwan	11	Thailand	1	Electrical and electronic equipment	872,455	1.30%	4. Untapped trade potential	.	.	5 to 10%
Taiwan	11	Thailand	4	Metal and metal products	293,988	2.20%	4. Untapped trade potential	.	.	15 to 20%
	13	Philippines		TOTAL	2,046,563	1.40%	4. Untapped trade potential
Taiwan	13	Philippines	1	Electrical and electronic equipment	896,268	1.30%	5. High untapped trade potential	.	.	0 to 5%
Taiwan	13	Philippines	4	Metal and metal products	142,082	1.00%	4. Untapped trade potential	.	.	0 to 5%
	19	Indonesia		TOTAL	1,303,826	0.90%	3. Predicted = Current or low values
Taiwan	19	Indonesia	5	Electrical and electronic equipment	104,530	0.20%	4. Untapped trade potential	.	.	0 to 5%

References

Anderson J.E., van Wincoop J.H. 2003,” Gravity with Gravitas : A Solution to the Boder Puzzle”, American Economic Review, vol. 93(1), pp. 170-92.

Bouet et al., 2001, “ Market Access Map: A Bilateral and Disaggregated Measure  of Market Access” Document de travail CEPII, 2001-18. Available at www.cepii.fr/anglaisgraph/workpap/summaries/2001/wp01-18.htm.

Chen, Kun-Ming, Ji Chou and Nai-Fong Kuo, 2004,” The Impact of East Asian Economic Integration on Taiwan”. Taiwan Economic Forum, October ,pp. 53-71.

Chen, Tain-Jy and Ying-Hua Ku, 2007,” Taiwan and East Asian Integration” in Peter Chow (ed) Economic Integration, Democratization and National Security in East Asia: Shifting Paradigm in US, China and Taiwan Relations”. Cheltenham.UK : Edward Elgar.pp. 172-190. 

Santos Silva J.M.C. and Tenreyro S., revised 2004, “The Log of Gravity”, FRB Boston Series, paper no. 03-1 (2003).

                                             

---

[1] Source: Taiwan Bureau of Foreign Trade, National Accounts, http://2k3dmz2.moea.gov.tw/GNWEB/english/indicators/reports/A.xls

[2] Source: Taiwan Bureau of Foreign Trade,, Economic Indicators http://2k3dmz2.moea.gov.tw/GNWEB/english/indicators/e_indicators.aspx?menu=2#sub02

[3] Annie Huang, “ROC, Honduras, El Salvador sign trilateral FTA”, Taiwan Journal, Vol XXIV, No.18.  Taiwan is also pursuing FTAs with the Dominican Republic and Costa Rica.

[4] Taiwan: EU official says any FTA will be ’some time’”, Taipei Times, April 20, 2007.

[5] Among those studies, Chen and Ku (2007) analyzed the impacts of ASEAN plus China, and ASEAN plus 3 on Taiwan’s economy. Chen  et al ( 2004) had 12 scenarios on various combinations of economic integration on Taiwan economy. 

[6] For a full description of the technical features of the model, see Hertel, T. W. (1997). Global Trade Analysis: Modeling and Applications, Cambridge: Cambridge University Press.

[7] The methodology for updating the protection data is that developed for the World Bank. For a description see Dominique van der Mensbrugghe, “Estimating the Benefits of Trade Reform: Why Numbers Change,” in World Bank, Trade, Doha, and Development: A Window into the Issues; http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/TRADE/0,,contentMDK:20732399~pagePK:148956~piPK:216618~theSitePK:239071,00.html; at p. 61.

[8] The comparative static version of the Linkage model produced income gains for industrialized countries under multilateral trade liberalization that were one third larger using the trade elasticities in the Linkage model compared to those in the GTAP 6.0 dataset. See Dominique van der Mensbrugghe, “Estimating the Benefits of Trade Reform: Why Numbers Change,” Chapter 4 in Trade, Doha, and Development: A Window into the Issues (World Bank; http://siteresources.worldbank.org/INTRANETTRADE/Resources/239054-1126812419270/4.EstimatingThe.pdf); at p. 71. 

[9] “Closure” in the context of a model refers to the choice of variables which are exogenous and which are endogenous.

[10] This is sometimes described as reflecting a medium-term time horizon in which labour supply is relatively “sticky.”

[11] The closure rule in which the rate of return to capital is fixed is sometimes described as reflecting longer-run “steady-state” growth conditions, For an example of the implications of fixing the return to capital and allowing investment to adjust, see John P. Gilbert, “GTAP Model Analysis: Simulating the Effect of a Korea-US FTA Using Computable General Equilibrium Techniques”; http://www.iie.com/publications/chapters_preview/326/appbiie311x.pdf. Gilbert reports net economic welfare gains for Korea that are 2.7 times larger, and for the U.S. that are 2.4 times larger, with this closure compared to standard closure.  For an example of the use of the labour market closure rule under which the wage rate is fixed, see Joseph F. Francois and Laura M. Baughman, “U.S.-Canadian Trade and U.S. State-Level Production and Employment”, in John M. Curtis and Dan Ciuriak (eds.) Trade Policy Research 2004 (Ottawa: DFAIT, 2004). `    

[12] See Gilbert (op. cit.) for a comparison of the impact of using alternative macroeconomic closures in the context of modelling the U.S.-Korea FTA. The fixed current account simulations substantially reduce the economic welfare gains for Korea (to 3/5 the level of the simulation with flexible current account) and marginally (by 5%) for the United States. 

[13] Chen and Ku (2007) Table 8.3 on p. 178.

[14] See unstats.un.org/unsd/cpmtrade.

[15] The application of MacMap since 2005 can be found at : www.macmap.org

[16] TradeSim uses this estimation technique to address the problem of zero trade levels.  

For more details, see the Appendix 4 of TradeSim ( third version). ITC working paper 2005. Geneva.

[17] “ The export potential is calculated on the basis of separate regressions for all 19 sectors,

applying equation [5] (the variable ij Geo was only applied to sector P1 and P2). For each

exporting country i and importing country j, we calculate the predicted values ijk Xˆ for sector

k, i.e. export potentials are equivalent to the within-sample predictions, which are obtained

through separate sector regressions. As we used fixed effects in the estimation, the application of an ‘a posteriori’ fixed effect Fi (as in TradeSim2) for the calculation of trade potentials in order to adjust the trade potential from systematic effects is not necessary. Hence, the trade potential TPijk is simply given by: ijk ijk TP = Xˆ .It is important to notice that the results obtained represent relative trade potentials, as the gravity specification employed is dependent on the sample specification. Hence, potentials should be understood as potentials within the specified model set-up, i.e. within the context of the specified 132 exporting and 154 importing countries, controlling for the variables included in [5] according to the sector analyzed, and the multilateral resistance terms. Any variation in the model or sample specification would necessarily entail different results”.

[18] Export potentials are limited by the gravity model which is constructed on a general equilibrium model; hence it does not capture the dynamic effects and cross-industry linkages. Secondly, the model is very sensitive to changes in the sample due to the inclusion of multilateral resistance terms. Thirdly, the impacts of FDI on export were ignored due to paucity of bilateral data of FDI and finally, trade complementarity between countries is not sufficiently taken into account in the gravity model.

[19] The top 15 sectors were ranked by the current export performances in 2002-2003. Obviously, the four sectors omitted were mostly in forestry / fishing products, mining/quarrying, petroleum and recycling.  See Appendix 1 for more details.